mlsanityData to 14 Oct 2026

Good Keypoints for the Two-View Geometry Estimation Problem

arXiv 2503.18767 ↗First posted 24 Mar 2025cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Local features are essential to many modern downstream applications. Therefore, it is of interest to determine the properties of local features that contribute to the downstream performance for a better design of feature detectors and descriptors. In our work, we propose a new theoretical model for scoring feature points (keypoints) in the context of the two-view geometry estimation problem. The model determines two properties that a good keypoint for solving the homography estimation problem should have: be repeatable and have a small expected measurement error. This result provides key insights into why maximizing the number of correspondences doesn't always lead to better homography estimation accuracy. We use the developed model to design a method that detects keypoints that benefit the homography estimation and introduce the Bounded NeSS-ST (BoNeSS-ST) keypoint detector. The novelty of BoNeSS-ST comes from strong theoretical foundations, a more accurate keypoint scoring due to subpixel refinement and a cost designed for superior robustness to low saliency keypoints. As a result, BoNeSS-ST outperforms prior self-supervised local feature detectors on the planar homography estimation task and is on par with them on the epipolar geometry estimation task.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 23 of its 41 references that are in the map. First posted on arXiv 24 Mar 2025.

Builds on

23 of 41 references.

References · 23

  1. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  2. 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,320
  3. 2017Deep Image Priorcited by 1,098
  4. 2015ORB-SLAM: a Versatile and Accurate Monocular SLAM Systemcited by 988
  5. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 984
  6. 2018MegaDepth: Learning Single-View Depth Prediction from Internet Photoscited by 737
  7. 2020DISK: Learning local features with policy gradientcited by 312
  8. 2023RoMa: Robust Dense Feature Matchingcited by 302
  9. 2017HPatches: A benchmark and evaluation of handcrafted and learned local descriptorscited by 264
  10. 2020Image Matching across Wide Baselines: From Paper to Practicecited by 225
  11. 2017Working hard to know your neighbor's margins: Local descriptor learning losscited by 212
  12. 2022DKM: Dense Kernelized Feature Matching for Geometry Estimationcited by 177
11 more
  1. 2017Learning to Find Good Correspondencescited by 176
  2. 2018LF-Net: Learning Local Features from Imagescited by 151
  3. 2020Patch2Pix: Epipolar-Guided Pixel-Level Correspondencescited by 129
  4. 2019Key.Net: Keypoint Detection by Handcrafted and Learned CNN Filterscited by 118
  5. 2021Learning Accurate Dense Correspondences and When to Trust Themcited by 113
  6. 2014TILDE: A Temporally Invariant Learned DEtectorcited by 55
  7. 2022TopicFM: Robust and Interpretable Topic-Assisted Feature Matchingcited by 32
  8. 2022Self-Supervised Equivariant Learning for Oriented Keypoint Detectioncited by 26
  9. 2023A Large Scale Homography Benchmarkcited by 15
  10. 2023NeSS-ST: Detecting Good and Stable Keypoints with a Neural Stability Score and the Shi-Tomasi Detectorcited by 5
  11. 2024GMM-IKRS: Gaussian Mixture Models for Interpretable Keypoint Refinement and Scoringcited by 5

Built on by

Not cited yet.